一种基于PCA的多模板字符识别

本文提出了一种基于主元分析法(PCA)的多模板字符识别算法.为了进一步提高数字和英文字符识别的鲁棒性,本文首先利用PCA算法把提取到的字符特征降维,再采用K-均值算法把每类字符聚类为多类,使同一类字符有多个模板,最后采用欧式距离实现多模板的匹配.本文将该算法用于巴西车牌字符识别,实验表明,该算法能有效地提高多字体字符的识别正确率,具有较高的实用价值.

Multi-template character recognition based on PCA

This paper presented a new multi-template character recognition method based on principal component analysis(PCA).To improve robust classification research on digital and English character recognition,this paper used PCA to dimension reduction of character features firstly,then used K-means to cluster each character to make each character have multiple templates,the last achieves multi-template match by Euclidean distance.This method has been used for Brazil car plate character recognition.Experiments show that this method can improve recognition accuracy of characters,and is of practical value obviously.